32bit vs 64bit - Why should I consider?

[StarsFan4Life]

Could the heat have anything to do with the fact that I am running dual 1900 x 1080 screens off of this one card?

 

[Gamingphreek]

Well your CPU's max temp isn't really the problem here - its your Video Card's temperature.

I don't know how long you have had the system, but try taking compressed air and cleaning out the heatsink and fan on the video card. Also try repositioning case fans. You may want to look at a new HSF for your video card.

Before we do that; there are ways to check. First off, keep your case open and try and keep the space around your video card free of anything. Run Futuremark or whatever benchmark utility you want and keep an eye on the temperatures. Try and keep it below 75C and see if it shuts down then (I doubt it will).

As for the dual monitors - well sure it contributes. Instead of having to render an image to fit on 1920x1080 pixels, it has to render double that. While you are at your desktop it really isn't a big issue; however, when you go into a game, you then have to render a TON more. Now disconnecting one of the screens might help temperatures, but the mere fact that your card reaches 75C in 30 seconds means your card is STARVED of cool air. If worse comes to worse, try and aim a wall fan at the card for now.

-Kevin

 

[StarsFan4Life]

Originally posted by: Gamingphreek
Well your CPU's max temp isn't really the problem here - its your Video Card's temperature.

I don't know how long you have had the system, but try taking compressed air and cleaning out the heatsink and fan on the video card. Also try repositioning case fans. You may want to look at a new HSF for your video card.

Before we do that; there are ways to check. First off, keep your case open and try and keep the space around your video card free of anything. Run Futuremark or whatever benchmark utility you want and keep an eye on the temperatures. Try and keep it below 75C and see if it shuts down then (I doubt it will).

As for the dual monitors - well sure it contributes. Instead of having to render an image to fit on 1920x1080 pixels, it has to render double that. While you are at your desktop it really isn't a big issue; however, when you go into a game, you then have to render a TON more. Now disconnecting one of the screens might help temperatures, but the mere fact that your card reaches 75C in 30 seconds means your card is STARVED of cool air. If worse comes to worse, try and aim a wall fan at the card for now.

-Kevin

Kevin,

This is a brand new build, about a week old. I will try the cooling suggestions. Question though, don't ATI's run hotter than NVidia cards?

 

[Gamingphreek]

Well that depends. The more powerful the GPU you get, generally speaking, the more power you use. Some power is then lost and dissipated as heat. So you would have to look at the power and thermal characteristics of a particular piece of hardware to get an idea of how much power it uses.

As of right now, I honestly have not been keeping up on Video Manf latest and greatest cards, so I couldn't offer and opinion on better options. My recommendation; however, is to stick with this one. If you can verify (I'm 90% sure) that it is overheating, then the solution is simple - more airflow and a better Heat-Sink Fan. If it begins to shutdown at other times, then it seems that the video card is no functioning correctly and you should seek an RMA.

What is the cooling around this video card like? Is the card up against another card given that, IIRC, it takes up 2 slots? Are the case fans both intake, exhaust, or a combination of the 2? Is the fan working on the video card? Is the heatsink hot (Note: It should be very hot if it is working correctly)?

-Kevin

 

[StarsFan4Life]

I will take pics tonight and post them. I did a very very clean cable install. The video card is the only thing I have in any of the slots. Although, I wonder if the southbridge heatsink has anything to do with this:

http://www.productwiki.com/upl...ma78gm_s2h-400-400.jpg

 

[Gamingphreek]

Originally posted by: StarsFan4Life
I will take pics tonight and post them. I did a very very clean cable install. The video card is the only thing I have in any of the slots. Although, I wonder if the southbridge heatsink has anything to do with this:

http://www.productwiki.com/upl...ma78gm_s2h-400-400.jpg

Southbridge generally contains mostly I/O controllers and additional PCI-E lanes. It generally doesn't get very hot at all which is why it is only a HS for cooling.

Is the fan indeed working on your 4870? When running GPU intensive apps like that, does the HS get warm? Can you hear the fan ramp up with more intense GPU base processing?

-Kevin

 

[StarsFan4Life]

Yes, I can hear the fan rev up. I believe it does get warm, but I can check tonight.

 

[StarsFan4Life]

Is turning off one of the monitors the same as unplugging it from the video card?

 

[Gamingphreek]

I don't believe turning it off is enough because I believe a monitor has a +5VSB (Volt StandBy). Unplugging it from the wall may do the trick though - though at that point it may be easier to simply unplug it from the computer.

If the heatsink is warm, that means it is making proper contact like it should.

-Kevin

 

[MrPickins]

75C sounds about right for a 4870 according to this:

http://www.tomshardware.com/re...n-hd-4870,1964-17.html

 

[Gamingphreek]

Originally posted by: MrPickins
75C sounds about right for a 4870 according to this:

http://www.tomshardware.com/re...n-hd-4870,1964-17.html

Jeez - that is stupidly hot. Whoever the chief architect was should be let go IMO!

If the screen goes blank after 30 seconds of GPU load and then 113 seconds of lighter GPU load though, it still has to be a problem with the GPU in my mind.

-Kevin

 

[MrPickins]

Originally posted by: Gamingphreek

Originally posted by: MrPickins
75C sounds about right for a 4870 according to this:

http://www.tomshardware.com/re...n-hd-4870,1964-17.html

Jeez - that is stupidly hot. Whoever the chief architect was should be let go IMO!

If the screen goes blank after 30 seconds of GPU load and then 113 seconds of lighter GPU load though, it still has to be a problem with the GPU in my mind.

-Kevin

http://forums.anandtech.com/me...y&keyword1=temperature

As long as failure rates are low, I don't see it as being a problem.

 

[Gamingphreek]

Originally posted by: MrPickins

Originally posted by: Gamingphreek

Originally posted by: MrPickins
75C sounds about right for a 4870 according to this:

http://www.tomshardware.com/re...n-hd-4870,1964-17.html

Jeez - that is stupidly hot. Whoever the chief architect was should be let go IMO!

If the screen goes blank after 30 seconds of GPU load and then 113 seconds of lighter GPU load though, it still has to be a problem with the GPU in my mind.

-Kevin

http://forums.anandtech.com/me...y&keyword1=temperature

As long as failure rates are low, I don't see it as being a problem.

Sure it isn't a problem but it is horrible design/engineering. That much heat means a ton of power is being used. Given that it is a 4870 and isn't the fastest thing on the planet, it shouldn't be wasting that much power (be it by Transistor Leakage or otherwise).

OP - I would still check and see when the Video Card shuts down (By when I mean at what temperature). If it is a consistent 75C, it could be a problem with the card's fan controller or something. If there is a way you can throw a test card in there instead and run Futuremark - you can really isolate whether it is the graphics card or not.

-Kevin

 

[MrPickins]

Originally posted by: Gamingphreek

Originally posted by: MrPickins
http://forums.anandtech.com/me...y&keyword1=temperature

As long as failure rates are low, I don't see it as being a problem.

Sure it isn't a problem but it is horrible design/engineering. That much heat means a ton of power is being used. Given that it is a 4870 and isn't the fastest thing on the planet, it shouldn't be wasting that much power (be it by Transistor Leakage or otherwise).

The 4870 is a 150W part capable of 1200 Gflops (according to wikipedia). Intel's i7 chips are limited to well under 100 Gflops.

GPU's do a shit-ton more work than they get credit for, and the power has to come from somewhere. There is no free lunch.


Back on topic: OP, do you have any spare video cards or ones you can borrow? I agree that it sounds like a hardware issue.

 

[Gamingphreek]

The 4870 is a 150W part capable of 1200 Gflops (according to wikipedia). Intel's i7 chips are limited to well under 100 Gflops.

GPU's do a shit-ton more work than they get credit for, and the power has to come from somewhere. There is no free lunch.

General Purpose vs Single Purpose

Sure the 4870 is capable of 1200Gflops in highly parallel vector and floating point code. A GPU however wouldn't be able to do much at all with the other tasks a CPU has to cope with.

I was more referring to the fact that the Nvidia equivalent processes just as much while using less power. I hate having a leaf blower (Ah fun to reminisce about the Geforce FX ) running wild in my case.

At any rate, like you said, I think we are resolved that there were, in fact, 2 problems the OP was having. 1 was software (Which is fixed) and the other is H/W (Not fixed).

-Kevin

 

[dguy6789]

75C is a perfectly normal and fine temperature for the 4870. Temperatures don't seem to be your problem. If it is crashing in furmark though, it's probable that your card is defective. See if you can get a replacement.

 

[MrPickins]

Originally posted by: Gamingphreek
I was more referring to the fact that the Nvidia equivalent processes just as much while using less power. I hate having a leaf blower (Ah fun to reminisce about the Geforce FX ) running wild in my case.

All of Nvidia's 200 series GPU's are 145w+ according to wikipedia.

Pushing that many triangles takes a lot of processing power.

 

[StarsFan4Life]

Originally posted by: Gamingphreek

Originally posted by: MrPickins

Originally posted by: Gamingphreek

Originally posted by: MrPickins
75C sounds about right for a 4870 according to this:

http://www.tomshardware.com/re...n-hd-4870,1964-17.html

Jeez - that is stupidly hot. Whoever the chief architect was should be let go IMO!

If the screen goes blank after 30 seconds of GPU load and then 113 seconds of lighter GPU load though, it still has to be a problem with the GPU in my mind.

-Kevin

http://forums.anandtech.com/me...y&keyword1=temperature

As long as failure rates are low, I don't see it as being a problem.

Sure it isn't a problem but it is horrible design/engineering. That much heat means a ton of power is being used. Given that it is a 4870 and isn't the fastest thing on the planet, it shouldn't be wasting that much power (be it by Transistor Leakage or otherwise).

OP - I would still check and see when the Video Card shuts down (By when I mean at what temperature). If it is a consistent 75C, it could be a problem with the card's fan controller or something. If there is a way you can throw a test card in there instead and run Futuremark - you can really isolate whether it is the graphics card or not.

-Kevin

Kevin,

I took the side panel off and have a large fan blowing full blast on the card. While the rest of the system stays cool, the card STILL heats up to 75c for about 66 seconds and then shuts off. The whole damn card (top and bottom) is hot as hell. I think I regret buying this card!

 

[StarsFan4Life]

Kevin,

I just took out my 4870, uninstalled it and put in my trusty ole EVGA 8600gts 512mb video card. At 1900 x 1080 running the stability test, I ran it for 5 minutes. It leveled out at 80c but didn't crash on me at all.

Are we safe to say that the 4870 is bad card?

 

[dguy6789]

If your machine passes prime 95 and memtest but fails furmark, it's pretty safe to say that the video card is bad. See if you can return it. Seems like you were unlucky and were shipped a defective card.

 

[Gamingphreek]

Yea it looks like your video card is bad. It also looks like you have isolated the problem to one of the specific power planes on your card because this happens at 75C only. So when the card goes to ramp the fan up further or increase clockspeeds or enable Unified Shader Pipelines, it crashes.

So simply RMA the card and get a new one - its no big deal, happens to everyone.

-Kevin

 

[StarsFan4Life]

Purchased through Newegg, RMA in progress. Hopefully this second one will be ok.

 

[Itchrelief]

My 4890 hits low 80s when Folding. 9800GTX+ hits Mid 70s. Under load, that's not an unusual temperature.

Unless there's a localized hotspot, I don't think the 75C temp is necessarily what's causing the crashes.

 

[StarsFan4Life]

Originally posted by: Itchrelief
My 4890 hits low 80s when Folding. 9800GTX+ hits Mid 70s. Under load, that's not an unusual temperature.

Unless there's a localized hotspot, I don't think the 75C temp is necessarily what's causing the crashes.

It has to be a hardware issue. Remember, I put in an EVGA 8600GTS and ran Furmark at the same resolution. The card got up past 80c and still kept going.

 

[F1shF4t]

Originally posted by: Gamingphreek
Sure it isn't a problem but it is horrible design/engineering. That much heat means a ton of power is being used. Given that it is a 4870 and isn't the fastest thing on the planet, it shouldn't be wasting that much power (be it by Transistor Leakage or otherwise).

-Kevin

Temperature is not equal to power consumption. Sticking a masive heatsink on it and limiting it to 30C will not make it consume significantly less power. (In reality there should be a slight difference due to semiconductor properties, if I remember correctly )

On the other hand the original 4870 512 card I had used to idle at 82C and load during heavy gaming to 86C. I don't see that as a bad thing, since the thermal stress on the chip will be much lower.

Also while a 4870 should be able to run Furmark, I wouldn't do it for extended periods of time as the vrm get insanely hot (my msi card stable but 120C on them, sapphire one doesn't have temp monitoring for them, it might be similar design to palit cards).